**3.1 Genetic modifications in bioreactor**

Genetic modification is one of the biggest steps in producing cell therapy products. In biologics manufacturing, it has been practiced for many years for producing antibodies, proteins, or other biotechnological drugs. It has also been used extensively in the cell therapy industry as various cell-based products have been applied for treating multiple incurable genetic diseases in recent years. Some genetic modifications affect patients directly and some indirectly. For example, in adrenoleukodystrophy (ALD), a neurological disorder occurs due to malfunction of oligodendrocytes and microglia where genetic modification can affect a patient directly. To recover from it, a corrected gene is inserted into the patient-derived hPSCs and transplanted into the

#### **Figure 2.**

*Schematic illustrations of integrated single-step cell manufacturing strategies in bioreactor culture for stem cell therapy applications. Skin cells are isolated from the patient and reprogrammed to hiPSCs on microcarriers using a nonviral approach. After expansion as aggregates, hiPSCs are stored in a master cell bank or differentiated directly in bioreactor. After performing characterization, quality assurance, and screening for safety and efficacy, cells are delivered to hospital or stored in a cell bank for future use.*

**75**

all the three germ layers [72].

using targeting nucleases [78–80].

*Integrated Biologics Manufacturing in Stirred-Suspension Bioreactor: A Stem Cell Perspective*

patient's brain, which is differentiated into microglia to promote production of myelin

In some gene therapy applications, patients are exposed indirectly to genetic modification. For example, in thalassemia, patient blood cells are extracted from the body and the cells are modified and enriched in *ex vivo* to target the specific antigens of patients' body [55]. Other indirect genetic modifications used for treating CAR or T-cell receptor (TCR) genes to T-cells [56], expression of CD40 ligand in dendritic cells [57], adenosine-deaminase severe-combined immunodeficiency [58], and betathalassemia [59], as well as deletion or insertion of desired genes in a specific genomic location. Among them, CAR-T cell therapy has got much attention for treating cancer-related diseases. These genetically modified T-cells can specifically target the antigens and kill the cancer cells efficiently [60]. CARs and TCRs are the mostly used receptors which are engineered to activate the T-cells [61]. Nowadays, a lot of CAR-T cell-based therapies are being established for treating advanced-stage lymphoma [62]

Viral vectors are commonly used to deliver genetic cargo to cells (**Figure 1**). This involves a two-step process: preparation for viral vectors and transduction for modifying the cells to express desired property. Lentiviral and gamma-retroviral are widely-used for their superior transduction efficiency but their transgenes are integrated with the host genome [65]. Another choice for viral transduction is adenovirus where viral transgenes are not integrated into the host genome but less efficient than lenti- and retro-virus. The major drawbacks in viral vector mediated transduction are concerns for safety of the products [66]. Viral vectors are widely

Other methods for cellular transduction use nonviral approaches, including nucleofection or electroporation, or liposome-mediated delivery of DNA or RNA into cells. Although DNA vectors are easy to scale-up, carry large-size DNA with less immunotoxicity, this process is less efficient than the viral transduction. There are some other methods for skipping the use of viral vectors which are also efficient in doing the transgene expression [67–69]. Hsu et al. reported successful transfection by using commercially available nonviral cationic reagents, for example, TransIT-3D, TransIT-2020, XtremeGENE 9, XtremeGENE HP, JetPrime, Lipofectamine 3000, and Effectene and compared their transfection efficiency [70]. Warren et al. reported efficient reprogramming of hiPSCs from various cell sources by using mRNA and differentiated the cells into three germ layers [71]. hiPSCs were also reprogrammed by using recombinant protein that also maintained

Since transgene possesses high risk of cancer-causing agents; therefore, removal

after transduction is highly desired. There are a few methods developed for the removal of these vectors. One of the methods is the piggyBac transposon system, which has been used to remove tandem Yamanaka reprograming genes Oct4, Sox2, Klf4, and c-Myc from iPSCs following reprogramming [73]. Removal of transgenes after incorporating CAR into T-cells used another transposon system called Sleeping Beauty, which successfully removed any genetic scar from the transduced cells [74, 75]. Likewise, transgene-free iPSCs have also been produced by Cre excision of reprogramming genes via loxP sites [76]. Integration-deficient viral vectors are also good candidates for producing transgene-free cell therapy products by mutating viral integrase [77]. Another approach is to use site-directed integration

Various genome engineering technologies have been explored for gene addition, deletion, or correction in the cell therapy industry and are increasing day by day [81]. The most widely used targeting nucleases are zinc-finger nucleases (ZFNs), clustered regulatory interspaced short palindromic repeats (CRISPR)/Cas

and B-cell lymphoma [63] as well as other autoimmune diseases [64].

used for reprogramming hiPSCs from skin fibroblasts cells [2].

*DOI: http://dx.doi.org/10.5772/intechopen.83813*

in the patient's brain that recovered the ALD [54].

#### *Integrated Biologics Manufacturing in Stirred-Suspension Bioreactor: A Stem Cell Perspective DOI: http://dx.doi.org/10.5772/intechopen.83813*

patient's brain, which is differentiated into microglia to promote production of myelin in the patient's brain that recovered the ALD [54].

In some gene therapy applications, patients are exposed indirectly to genetic modification. For example, in thalassemia, patient blood cells are extracted from the body and the cells are modified and enriched in *ex vivo* to target the specific antigens of patients' body [55]. Other indirect genetic modifications used for treating CAR or T-cell receptor (TCR) genes to T-cells [56], expression of CD40 ligand in dendritic cells [57], adenosine-deaminase severe-combined immunodeficiency [58], and betathalassemia [59], as well as deletion or insertion of desired genes in a specific genomic location. Among them, CAR-T cell therapy has got much attention for treating cancer-related diseases. These genetically modified T-cells can specifically target the antigens and kill the cancer cells efficiently [60]. CARs and TCRs are the mostly used receptors which are engineered to activate the T-cells [61]. Nowadays, a lot of CAR-T cell-based therapies are being established for treating advanced-stage lymphoma [62] and B-cell lymphoma [63] as well as other autoimmune diseases [64].

Viral vectors are commonly used to deliver genetic cargo to cells (**Figure 1**). This involves a two-step process: preparation for viral vectors and transduction for modifying the cells to express desired property. Lentiviral and gamma-retroviral are widely-used for their superior transduction efficiency but their transgenes are integrated with the host genome [65]. Another choice for viral transduction is adenovirus where viral transgenes are not integrated into the host genome but less efficient than lenti- and retro-virus. The major drawbacks in viral vector mediated transduction are concerns for safety of the products [66]. Viral vectors are widely used for reprogramming hiPSCs from skin fibroblasts cells [2].

Other methods for cellular transduction use nonviral approaches, including nucleofection or electroporation, or liposome-mediated delivery of DNA or RNA into cells. Although DNA vectors are easy to scale-up, carry large-size DNA with less immunotoxicity, this process is less efficient than the viral transduction. There are some other methods for skipping the use of viral vectors which are also efficient in doing the transgene expression [67–69]. Hsu et al. reported successful transfection by using commercially available nonviral cationic reagents, for example, TransIT-3D, TransIT-2020, XtremeGENE 9, XtremeGENE HP, JetPrime, Lipofectamine 3000, and Effectene and compared their transfection efficiency [70]. Warren et al. reported efficient reprogramming of hiPSCs from various cell sources by using mRNA and differentiated the cells into three germ layers [71]. hiPSCs were also reprogrammed by using recombinant protein that also maintained all the three germ layers [72].

Since transgene possesses high risk of cancer-causing agents; therefore, removal after transduction is highly desired. There are a few methods developed for the removal of these vectors. One of the methods is the piggyBac transposon system, which has been used to remove tandem Yamanaka reprograming genes Oct4, Sox2, Klf4, and c-Myc from iPSCs following reprogramming [73]. Removal of transgenes after incorporating CAR into T-cells used another transposon system called Sleeping Beauty, which successfully removed any genetic scar from the transduced cells [74, 75]. Likewise, transgene-free iPSCs have also been produced by Cre excision of reprogramming genes via loxP sites [76]. Integration-deficient viral vectors are also good candidates for producing transgene-free cell therapy products by mutating viral integrase [77]. Another approach is to use site-directed integration using targeting nucleases [78–80].

Various genome engineering technologies have been explored for gene addition, deletion, or correction in the cell therapy industry and are increasing day by day [81]. The most widely used targeting nucleases are zinc-finger nucleases (ZFNs), clustered regulatory interspaced short palindromic repeats (CRISPR)/Cas

*Current Topics in Biochemical Engineering*

the conventional batch processing.

**3.1 Genetic modifications in bioreactor**

single-step approach for cell-based therapies (**Figure 2**).

they respectively reported successful production of monoclonal antibody as well as highly complex, less stable pharmaceutical protein with consistent product quality, high product output, and low cost. Process integrity is necessary for reducing cumbersome production steps and cutting cost significantly. One such integrated system developed by Johnson & Johnson has recently got FDA approval for largescale HIV drug production [53] that reduces time and cost by one third compared to

Since biologics production for cell therapies require multiple steps, integration of all of the steps will give high product quality and safety, as well as help overcome stringent regulatory requirements. In this context, we will discuss how to integrate some important basic steps of cell manufacturing especially genetic modification, cellular reprogramming, expansion, and differentiation in bioreactors to promote a

Genetic modification is one of the biggest steps in producing cell therapy products. In biologics manufacturing, it has been practiced for many years for producing antibodies, proteins, or other biotechnological drugs. It has also been used extensively in the cell therapy industry as various cell-based products have been applied for treating multiple incurable genetic diseases in recent years. Some genetic modifications affect patients directly and some indirectly. For example, in adrenoleukodystrophy (ALD), a neurological disorder occurs due to malfunction of oligodendrocytes and microglia where genetic modification can affect a patient directly. To recover from it, a corrected gene is inserted into the patient-derived hPSCs and transplanted into the

*Schematic illustrations of integrated single-step cell manufacturing strategies in bioreactor culture for stem cell therapy applications. Skin cells are isolated from the patient and reprogrammed to hiPSCs on microcarriers using a nonviral approach. After expansion as aggregates, hiPSCs are stored in a master cell bank or differentiated directly in bioreactor. After performing characterization, quality assurance, and screening for* 

*safety and efficacy, cells are delivered to hospital or stored in a cell bank for future use.*

**74**

**Figure 2.**

endonucleases, or transcription-activator like effector nucleases (TALENs) [82]. Although the CRISPR/Cas system has recently received much attention due to broad use in genome engineering of patient cells [83], ZFNs are also popular for treating graft-versus-host disease in T-cell therapy [84].

Recently, a nuclease dead variant of Cas9 bearing a transcriptional trans-activator has recently been used in cellular reprogramming by activating the transcription factors Oct4 and Sox2, which maintained pluripotency and expressed the markers for the three germ layers [85].

Although genetic modification is a rate-limiting step in the cell manufacturing industry, the conventional methods make it more complicated because it is a multistep process. Conventional genetic modification in planar culture is also costly, labor-intensive, and time-consuming. The bioreactor is a better platform for producing large-scale genetically modified cells for commercial purposes because cell expansion is possible in the same vessel which makes the process straightforward (**Figure 2**). For genetic modification in bioreactor, Hsu et al. recently reported how to transfect reprogramming factors in bioreactor where they tried eliminating viral vectors for gene delivery by using cationic reagents [78]. Generally, transfection of reprogramming factors for generating induced pluripotent stem cells (iPSCs) is done in adherent culture and then cells are expanded in 2D or 3D which is a twostep process. By integrating the genetic modification step in bioreactor, it is possible to establish a single-step process which enables cell manufacturing in automated and closed bioreactor system.

Genetic modification is also a challenging step in CAR-T cell therapy-based products. In CAR-T cell therapy, generally cells are isolated from patients' blood sample and then the cells are expanded after selection and activation. Finally, the cells need to be transduced with the CAR or any other antigens depending on target diseases. Conventional methods for genetic transduction are based on planar culture where every step is performed in open culture system. Recently, a few steps are integrated in bag culture system where selection, activation, and expansion can be done in a single step using DynaMag™ CTS™ [86], whereas the Xuri cell expansion System developed by GE Healthcare can expand cells in large numbers [87–89].

Although washing and concentrating the final product are integrated by the COBE® 2991 system developed by Terumo BCT [90], the transduction step is still not integrated in any of the above systems. Integrating the transduction step with the expansion and formulation will make the CAR-T cell therapy straightforward and performing these steps in bioreactor is a good platform since the physiological parameters as well as automated operation is possible in bioreactor culture. Miltenyi Biotec developed a device named CliniMACS Prodigy™ which is based on bag culture for CAR-T cell therapy. This device integrated major steps especially cell preparation, selection, activation, expansion, transduction, washing, and formulation in an automated system [91–93]. Such integration in the bioreactor will pave a straightforward method for producing cell-based products in a closed and automatic method.

#### **3.2 Integrated system for large-scale expansion and differentiation in bioreactor**

Current manufacturing practices for stem cell-based products are multi-step: derivation, expansion, and differentiation. In this process, patient-derived skin fibroblast cells are transduced with reprogramming factors in the planar culture. After deriving hiPSCs, cells are expanded in planar or bioreactor culture to obtain a large number of cells. Then cells are differentiated to target cells of interest. The differentiated cells are characterized and transplanted to the patient in a dosedependent manner. As this process is complicated with multiple steps, it poses high risk of contamination to the final products. Moreover, maintaining cGMP culture

**77**

*Integrated Biologics Manufacturing in Stirred-Suspension Bioreactor: A Stem Cell Perspective*

Here, the bioreactor may be a good platform for doing this (**Figure 2**).

The bioreactor platform is widely used for the large-scale expansion of hPSC-based cell therapy production because bioreactor is easy to operate in an automated mode where various physicochemical parameters can be regulated in a closed-system. Two groups have demonstrated that the bioreactor is conducive to cellular reprogramming [97, 98]. Shafa et al. reported a significantly higher reprogramming efficiency in the bioreactor compared to the planar culture [97]. Since mesenchymal-epithelial transition (MET) is an important early step in cellular reprogramming [99], transformed fibroblasts that are moved into the bioreactor will form aggregates that are efficiently expanded in the bioreactor. Indeed because fibroblasts are substrate-dependent, bioreactor culture may be promoting aggregate

platform is also mandatory for cell-therapy products [94–96], which makes the cell manufacturing process more complicated. Therefore, developing an integrated system that can combine all these steps from derivation to final products is required.

Unfortunately, bioreactor reprogramming methods require genetic modification (retroviral, piggyBAC) prior to bioreactor expansion. It is theoretically possible to pursue cellular reprograming fully and completely in the bioreactor. Recently, for example, Hsu et al. has demonstrated that it is possible to transfect human fibroblasts directly on microcarriers [70]. Reprogrammed cells in theory will leave the

Following bioreactor derivation of hPSCs, the next big steps are expansion and differentiation. Generally, a large number of cells are required for an effective cell

In the conventional process, cell expansion is performed in planar culture. However, it has many drawbacks and limits the cell expansion in various ways. Planar culture

Another major drawback is surface coating. Extracellular matrix (ECM) is needed for surface coating which is initially derived from animal sources, which poses high risk in clinical-grade manufacturing. Currently, recombinant ECM has been discovered, which can be used efficiently for clinical applications [101]. The advancement in cell coating also stimulated the advancement of integration and

Automated planar culture systems have been established for the expansion of hPSCs for clinical-grade cell manufacturing. One of the notable automated systems for cell manufacturing is CompacT SelecT™ developed by the TAP Biosystems. This system is based on T-flask where 90 T175 flasks can be accommodated for large-scale expansion of cells. All the cell culture steps, cell counting, seeding, medium change, passaging, and plating as well as transient transfection can be done automatically by using this robotic system. However, such systems are not used for differentiation since differentiation is a complicated process, which needs several components to add in the culture medium. As a result, the expansion and differen-

Cell expansion in bioreactors need not require surface coating except for microcarrier culture. Bioreactor also provides enough growth surface availability. Generally, a single bioreactor (100 mL working volume) is enough for providing clinically relevant number of cells for autologous cell therapy applications. Several types of bioreactors are employed for the expansion of hPSCs [102]. For anchoragedependent expansion of hPSCs, microcarriers need to be coated with ECM for cell

After large-scale expansion, cells are harvested by detaching them from the microcarrier using enzymatic treatment. Nonenzymatic detachment is also available by changing temperature or pH [18–20]. Bioreactor expansion of hPSCs on

is unable to provide enough growth surfaces for the unlimited expansion.

to 1010 cells per 70 kg patient [100].

*DOI: http://dx.doi.org/10.5772/intechopen.83813*

formation and therefore cellular reprogramming.

therapy application, which is ranging from 108

automation of cell expansion in adherent culture.

tiation process in planar culture is mostly disintegrated.

attachment in the bioreactor [100, 103–105].

microcarrier to form aggregates in the bioreactor via MET.

#### *Integrated Biologics Manufacturing in Stirred-Suspension Bioreactor: A Stem Cell Perspective DOI: http://dx.doi.org/10.5772/intechopen.83813*

platform is also mandatory for cell-therapy products [94–96], which makes the cell manufacturing process more complicated. Therefore, developing an integrated system that can combine all these steps from derivation to final products is required. Here, the bioreactor may be a good platform for doing this (**Figure 2**).

The bioreactor platform is widely used for the large-scale expansion of hPSC-based cell therapy production because bioreactor is easy to operate in an automated mode where various physicochemical parameters can be regulated in a closed-system. Two groups have demonstrated that the bioreactor is conducive to cellular reprogramming [97, 98]. Shafa et al. reported a significantly higher reprogramming efficiency in the bioreactor compared to the planar culture [97]. Since mesenchymal-epithelial transition (MET) is an important early step in cellular reprogramming [99], transformed fibroblasts that are moved into the bioreactor will form aggregates that are efficiently expanded in the bioreactor. Indeed because fibroblasts are substrate-dependent, bioreactor culture may be promoting aggregate formation and therefore cellular reprogramming.

Unfortunately, bioreactor reprogramming methods require genetic modification (retroviral, piggyBAC) prior to bioreactor expansion. It is theoretically possible to pursue cellular reprograming fully and completely in the bioreactor. Recently, for example, Hsu et al. has demonstrated that it is possible to transfect human fibroblasts directly on microcarriers [70]. Reprogrammed cells in theory will leave the microcarrier to form aggregates in the bioreactor via MET.

Following bioreactor derivation of hPSCs, the next big steps are expansion and differentiation. Generally, a large number of cells are required for an effective cell therapy application, which is ranging from 108 to 1010 cells per 70 kg patient [100]. In the conventional process, cell expansion is performed in planar culture. However, it has many drawbacks and limits the cell expansion in various ways. Planar culture is unable to provide enough growth surfaces for the unlimited expansion.

Another major drawback is surface coating. Extracellular matrix (ECM) is needed for surface coating which is initially derived from animal sources, which poses high risk in clinical-grade manufacturing. Currently, recombinant ECM has been discovered, which can be used efficiently for clinical applications [101]. The advancement in cell coating also stimulated the advancement of integration and automation of cell expansion in adherent culture.

Automated planar culture systems have been established for the expansion of hPSCs for clinical-grade cell manufacturing. One of the notable automated systems for cell manufacturing is CompacT SelecT™ developed by the TAP Biosystems. This system is based on T-flask where 90 T175 flasks can be accommodated for large-scale expansion of cells. All the cell culture steps, cell counting, seeding, medium change, passaging, and plating as well as transient transfection can be done automatically by using this robotic system. However, such systems are not used for differentiation since differentiation is a complicated process, which needs several components to add in the culture medium. As a result, the expansion and differentiation process in planar culture is mostly disintegrated.

Cell expansion in bioreactors need not require surface coating except for microcarrier culture. Bioreactor also provides enough growth surface availability. Generally, a single bioreactor (100 mL working volume) is enough for providing clinically relevant number of cells for autologous cell therapy applications. Several types of bioreactors are employed for the expansion of hPSCs [102]. For anchoragedependent expansion of hPSCs, microcarriers need to be coated with ECM for cell attachment in the bioreactor [100, 103–105].

After large-scale expansion, cells are harvested by detaching them from the microcarrier using enzymatic treatment. Nonenzymatic detachment is also available by changing temperature or pH [18–20]. Bioreactor expansion of hPSCs on

*Current Topics in Biochemical Engineering*

the three germ layers [85].

and closed bioreactor system.

graft-versus-host disease in T-cell therapy [84].

endonucleases, or transcription-activator like effector nucleases (TALENs) [82]. Although the CRISPR/Cas system has recently received much attention due to broad use in genome engineering of patient cells [83], ZFNs are also popular for treating

has recently been used in cellular reprogramming by activating the transcription factors Oct4 and Sox2, which maintained pluripotency and expressed the markers for

Recently, a nuclease dead variant of Cas9 bearing a transcriptional trans-activator

Although genetic modification is a rate-limiting step in the cell manufacturing industry, the conventional methods make it more complicated because it is a multistep process. Conventional genetic modification in planar culture is also costly, labor-intensive, and time-consuming. The bioreactor is a better platform for producing large-scale genetically modified cells for commercial purposes because cell expansion is possible in the same vessel which makes the process straightforward (**Figure 2**). For genetic modification in bioreactor, Hsu et al. recently reported how to transfect reprogramming factors in bioreactor where they tried eliminating viral vectors for gene delivery by using cationic reagents [78]. Generally, transfection of reprogramming factors for generating induced pluripotent stem cells (iPSCs) is done in adherent culture and then cells are expanded in 2D or 3D which is a twostep process. By integrating the genetic modification step in bioreactor, it is possible to establish a single-step process which enables cell manufacturing in automated

Genetic modification is also a challenging step in CAR-T cell therapy-based products. In CAR-T cell therapy, generally cells are isolated from patients' blood sample and then the cells are expanded after selection and activation. Finally, the cells need to be transduced with the CAR or any other antigens depending on target diseases. Conventional methods for genetic transduction are based on planar culture where every step is performed in open culture system. Recently, a few steps are integrated in bag culture system where selection, activation, and expansion can be done in a single step using DynaMag™ CTS™ [86], whereas the Xuri cell expansion System developed by GE Healthcare can expand cells in large numbers [87–89]. Although washing and concentrating the final product are integrated by the COBE® 2991 system developed by Terumo BCT [90], the transduction step is still not integrated in any of the above systems. Integrating the transduction step with the expansion and formulation will make the CAR-T cell therapy straightforward and performing these steps in bioreactor is a good platform since the physiological parameters as well as automated operation is possible in bioreactor culture. Miltenyi Biotec developed a device named CliniMACS Prodigy™ which is based on bag culture for CAR-T cell therapy. This device integrated major steps especially cell preparation, selection, activation, expansion, transduction, washing, and formulation in an automated system [91–93]. Such integration in the bioreactor will pave a straightforward

method for producing cell-based products in a closed and automatic method.

**3.2 Integrated system for large-scale expansion and differentiation in bioreactor**

Current manufacturing practices for stem cell-based products are multi-step: derivation, expansion, and differentiation. In this process, patient-derived skin fibroblast cells are transduced with reprogramming factors in the planar culture. After deriving hiPSCs, cells are expanded in planar or bioreactor culture to obtain a large number of cells. Then cells are differentiated to target cells of interest. The differentiated cells are characterized and transplanted to the patient in a dosedependent manner. As this process is complicated with multiple steps, it poses high risk of contamination to the final products. Moreover, maintaining cGMP culture

**76**

microcarrier is troublesome for clinical application because it needs an extra step for microcarrier separation from the final cell harvest. On the other hand, aggregate culture in bioreactors may not necessarily need a detachment step for harvesting [21–25] and clinically relevant numbers of cells can be produced in a single bioreactor as aggregate [21, 106–108].

A major drawback in aggregate culture is the size limitation. With the increase in aggregate size, the growth potential decreases in the large size aggregate due to diffusion limitation of oxygen and nutrients [109]. Therefore, maintaining aggregate size is an important issue to maintain high growth rate as well as high quality for cell therapy applications [21].

After expansion, cells can be differentiated in the same vessel which makes bioreactor culture a unique choice for integrated biologics manufacturing. Bioreactors were used for differentiation of hPSCs into various cell types, especially for cardiac [110–112], hepatic [113, 114], and neural [115] lineages. To provide straightforward methods for clinical applications, integration of expansion and differentiation is important and there are several reports published recently where expansion and differentiation were integrated [108, 116–118]. However, the integration of derivation with expansion and differentiation is still facing complications and there are a very few reports available.

Steiner et al. reported integration of derivation, propagation and differentiation of hESCs in suspension culture where hESCs were isolated from the inner cell mass in suspension culture that did not involve feeder cells or microcarriers [119]. However, the integration of derivation, expansion, and differentiation is not still realized for personalized medicine especially for autologous or allogenic cell therapy applications. Such integration is needed for overcoming the multi-step cell processing, which will reduce the risk of contamination and save cell processing time as well as reduce manufacturing costs for cell therapy manufacturing.
